This adds a .bdrv_open option to specify the external data file node.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

For external data files, data clusters must be excluded from the
refcount calculations. Instead, an implicit refcount of 1 is assumed for
the COPIED flag.

Compressed clusters and internal snapshots are incompatible with
external data files, so print an error if they are in use for images
with an external data file.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Internal snapshots and an external data file are incompatible because
snapshots require refcounting and non-linear mapping. Return an error
for all of the snapshot operations if an external data file is in use.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

This changes the qcow2 implementation to direct all guest data I/O to
s->data_file rather than bs->file, while metadata I/O still uses
bs->file. At the moment, this is still always the same, but soon we'll
add options to set s->data_file to an external data file.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Offset 0 cannot be assumed to mean an unallocated cluster any more.
Instead, the cluster type needs to be checked.

*file must refer to the data file instead of the image file if a valid
offset is returned from qcow2_co_block_status().
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

qcow2_alloc_compressed_cluster_offset() used to return the cluster
offset for success and 0 for error. This doesn't only conflict with 0 as
a valid host offset, but also loses the error code.

Similar to the change made to qcow2_alloc_cluster_offset() for
uncompressed clusters in commit 148da7ea, make the function return
0/-errno and return the allocated cluster offset in a by-reference
parameter.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

The cluster allocation code uses 0 as an invalid offset that is used in
case of errors or as "offset not yet determined". With external data
files, a host cluster offset of 0 becomes valid, though.

Define a constant INV_OFFSET (which is not cluster aligned and will
therefore never be a valid offset) that can be used for such purposes.

This removes the additional host_offset == 0 check that commit
ff52aab2 introduced; the confusion between an invalid offset and
(erroneous) allocation at offset 0 is removed with this change.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Offset 0 can be valid for normal (allocated) clusters now, so use
qcow2_get_cluster_type() instead.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Signed-off-by: NKevin Wolf <kwolf@redhat.com>

This adds basic constants, struct fields and helper function for
external data file support to the implementation.

QCOW2_INCOMPAT_MASK and QCOW2_AUTOCLEAR_MASK are not updated yet so that
opening images with an external data file still fails (we don't handle
them correctly yet).
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

This adds external data file to the qcow2 spec as a new incompatible
feature.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Image creation already involves a bdrv_co_truncate() call, which allows
to specify a preallocation mode. Just pass the right mode there and
remove the code that is made redundant by this.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>

Various sed regexp from common.filter use sed GNU extensions.
Instead of spending time to write these regex to be POSIX compliant,
verify the GNU sed is available and use it.
Signed-off-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Bash is not always installed as /bin/bash. In particular on OpenBSD,
the package installs it in /usr/local/bin.
Use the 'env' shebang to search bash in the $PATH.

Patch created mechanically by running:

  $ git grep -lE '#! ?/bin/bash' -- tests/qemu-iotests \
    | while read f; do \
      sed -i 's|^#!.\?/bin/bash$|#!/usr/bin/env bash|' $f; \
    done
Reviewed-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Bash is not always installed as /bin/bash. In particular on OpenBSD,
the package installs it in /usr/local/bin.
Use the 'env' shebang to search bash in the $PATH.
Reviewed-by: NKamil Rytarowski <n54@gmx.com>
Reviewed-by: NIgor Mammedov <imammedo@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: NThomas Huth <thuth@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Bash is not always installed as /bin/bash. In particular on OpenBSD,
the package installs it in /usr/local/bin.
Use the 'env' shebang to search bash in the $PATH.
Reviewed-by: NKamil Rytarowski <n54@gmx.com>
Reviewed-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Some test cases require specific formats. The method decorator
skip_if_unsupported() checks if requested formats are whitelisted.
The test #139 was selected for a sample output, after running
$ ./check -qcow2 131-140

137 3s ...
138 0s ...
139 2s ...
    [case not run] testBlkDebug (__main__.TestBlockdevDel): formats ['blkdebug'] are not whitelisted
    [case not run] testBlkVerify (__main__.TestBlockdevDel): formats ['blkverify'] are not whitelisted
    [case not run] testQuorum (__main__.TestBlockdevDel): formats ['quorum'] are not whitelisted
140 0s ...
Not run: 131 135 136
Some cases not run in: 139
Passed all 7 tests
Signed-off-by: NAndrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Supported formats listed by 'qemu' may differ from those listed by
'qemu-img' due to whitelists. Some test cases require specific formats
that may be used with qemu. They can be inquired directly by running
'qemu -drive format=help'. The response takes whitelists into account.
The method supported_formats() serves for that. The method decorator
skip_if_unsupported() checks if all requested formats are whitelisted.
If not, the test case will be skipped. That has been implemented in
the 'check' file in the way similar to the 'test notrun' mechanism.
Suggested-by: NRoman Kagan <rkagan@virtuozzo.com>
Suggested-by: NVladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: NKevin Wolf <kwolf@redhat.com>
Signed-off-by: NAndrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

bdrv_iterate_format (which is currently only used for printing out the
formats supported by the block layer) doesn't take format whitelisting
into account.

This creates a problem for tests: they enumerate supported formats to
decide which tests to enable, but then discover that QEMU doesn't let
them actually use some of those formats.

To avoid that, exclude formats that are not whitelisted from
enumeration, if whitelisting is in use.  Since we have separate
whitelists for r/w and r/o, take this a parameter to
bdrv_iterate_format, and print two lists of supported formats (r/w and
r/o) in main qemu.
Signed-off-by: NRoman Kagan <rkagan@virtuozzo.com>
Signed-off-by: NAndrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Replace the binary mode with the default text one when *.notrun
files are opened for skipped tests. That change is made for the
compatibility with Python 3 which returns error otherwise.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Signed-off-by: NAndrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

QEMU 2.12 (commit 1221fe6f) introduced
a new setting called l2-cache-entry-size that allows making entries on
the qcow2 L2 cache smaller than the cluster size.

I have been performing several tests with different cluster and entry
sizes and all of them show that reducing the entry size (aka L2 slice)
consistently improves I/O performance, notably during random I/O (all
tests done with sequential I/O show similar results). This is to be
expected because loading and evicting an L2 slice is more expensive
the larger the slice is.

Here are some numbers on fully populated 40GB qcow2 images. The
rightmost column represents the maximum L2 cache size in both cases.

   Cluster size = 64 KB
   |-------------+--------------+--------------+--------------|
   |             | 1MB L2 cache | 3MB L2 cache | 5MB L2 cache |
   |-------------+--------------+--------------+--------------|
   |  4KB slices |    6545 IOPS |   12045 IOPS |   55680 IOPS |
   | 16KB slices |    5177 IOPS |    9798 IOPS |   56278 IOPS |
   | 64KB slices |    2718 IOPS |    5326 IOPS |   57355 IOPS |
   |-------------+--------------+--------------+--------------|

   Cluster size = 256 KB
   |--------------+----------------+--------------+-----------------|
   |              | 512KB L2 cache | 1MB L2 cache | 1280KB L2 cache |
   |--------------+----------------+--------------+-----------------|
   |   4KB slices |      8539 IOPS |   21071 IOPS |      55417 IOPS |
   |  64KB slices |      3598 IOPS |    9772 IOPS |      57687 IOPS |
   | 256KB slices |      1415 IOPS |    4120 IOPS |      58001 IOPS |
   |--------------+----------------+--------------+-----------------|

As can be seen in the numbers, the only exception to the rule is when
the cache is large enough to hold all L2 tables. This is also to be
expected because in this case no cache entry is ever evicted so
reducing its size doesn't bring any benefit.

This patch sets the default L2 cache entry size to 4KB except when the
cache is large enough for the whole disk.
Signed-off-by: NAlberto Garcia <berto@igalia.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Test 238 does not require the kvm accelerator.  Using the qtest
accelerator allows the test to run in both non-kvm and non-tcg
environments.

iotests.VM implicitly uses the qtest accelerator and is really the class
that this test should be using.  Switch to that instead of
qemu.QEMUMachine.
Suggested-by: NThomas Huth <thuth@redhat.com>
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

- Update netlink types to linux v5.0
- fix accept4/getpeername/getsockname/recvfrom/recvmsg/read
- add/fix ELF_PLATFORM ofr aarch64 and arm
- fix "may be used uninitialized" warnings
- Fix breakpoint support in Nios
- Nicer strace output of chroot() syscall

# gpg: Signature made Thu 07 Mar 2019 10:06:16 GMT
# gpg:                using RSA key F30C38BD3F2FBE3C
# gpg: Good signature from "Laurent Vivier <lvivier@redhat.com>" [full]
# gpg:                 aka "Laurent Vivier <laurent@vivier.eu>" [full]
# gpg:                 aka "Laurent Vivier (Red Hat) <lvivier@redhat.com>" [full]
# Primary key fingerprint: CD2F 75DD C8E3 A4DC 2E4F  5173 F30C 38BD 3F2F BE3C

* remotes/vivier2/tags/linux-user-for-4.0-pull-request:
  linux-user: add new netlink types
  linux-user: Nicer strace output of chroot() syscall
  linux-user: fix "may be used uninitialized" warnings
  linux-user: don't short-circuit read with zero length
  Fix breakpoint support in Nios II user-mode emulation.
  linux-user: fix emulation of accept4/getpeername/getsockname/recvfrom syscalls
  linux-user: Fix ELF_PLATFORM for aarch64_be-linux-user
  linux-user: Add ELF_PLATFORM for arm
  linux-user: fix recvmsg emulation
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Python queue, 2019-02-22

Python:
* introduce "python" directory with module namespace
* log QEMU launch command line on qemu.QEMUMachine

Acceptance Tests:
* initrd 4GiB+ test
* migration test
* multi vm support in test class
* bump Avocado version and drop "🥑 enable"

# gpg: Signature made Fri 22 Feb 2019 19:37:07 GMT
# gpg:                using RSA key 657E8D33A5F209F3
# gpg: Good signature from "Cleber Rosa <crosa@redhat.com>" [marginal]
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg:          It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 7ABB 96EB 8B46 B94D 5E0F  E9BB 657E 8D33 A5F2 09F3

* remotes/cleber/tags/python-next-pull-request:
  Acceptance tests: expect boot to extract 2GiB+ initrd with linux-v4.16
  Acceptance tests: use linux-3.6 and set vm memory to 4GiB
  tests.acceptance: adds simple migration test
  tests.acceptance: adds multi vm capability for acceptance tests
  scripts/qemu.py: log QEMU launch command line
  Introduce a Python module structure
  Acceptance tests: drop usage of "🥑 enable"
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Enable building and installing rST docs with Sphinx

# gpg: Signature made Thu 07 Mar 2019 15:05:12 GMT
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [ultimate]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* remotes/pmaydell/tags/pull-sphinx-20190307:
  MAINTAINERS: Add entry for Sphinx documentation infrastructure
  docs/conf.py: Don't hard-code QEMU version
  Makefile: Abstract out "identify the pkgversion" code
  Makefile, configure: Support building rST documentation
  docs: Provide separate conf.py for each manual we want
  docs/conf.py: Disable option warnings
  docs/conf.py: Don't include rST sources in HTML build
  docs/conf.py: Configure the 'alabaster' theme
  docs/conf.py: Disable unused _static directory
  docs: Commit initial files from sphinx-quickstart
  docs: Convert memory.txt to rst format
  docs/cpu-hotplug.rst: Fix rST markup issues
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Add a MAINTAINERS entry for Sphinx documentation infrastructure:
this doesn't cover actual content, only the machinery we use to
build the docs.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-13-peter.maydell@linaro.org

Don't hard-code the QEMU version number into conf.py. Instead
we either pass it to sphinx-build on the command line, or
(if doing a standalone Sphinx run in a readthedocs.org setup)
extract it from the VERSION file.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Reviewed-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-12-peter.maydell@linaro.org
Message-id: 20190228145624.24885-12-peter.maydell@linaro.org

Abstract out the "identify the pkgversion" code from the
rule for creating qemu-version.h, so it sets makefile
variables for QEMU_PKGVERSION and QEMU_FULL_VERSION.
(We will want to use these when building the Sphinx docs.)

NB: As we abstract this out, we use -e to check for .git
rather than -d, since in some situations .git may be a file
rather than a directory.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-11-peter.maydell@linaro.org
Message-id: 20190228145624.24885-11-peter.maydell@linaro.org

Add support to our configure and makefile machinery for building
our rST docs into HTML files.

Building the documentation now requires that sphinx-build is
available; this seems better than allowing half the docs to
be built if it is not present but having half of them missing.
(In particular it means that assuming that distros configured with
--enable-docs they'll get a helpful error from configure telling
them the new build dependency.)
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-10-peter.maydell@linaro.org
Message-id: 20190228145624.24885-10-peter.maydell@linaro.org

By default Sphinx wants to build a single manual at once.
For QEMU, this doesn't suit us, because we want to have
separate manuals for "Developer's Guide", "User Manual",
and so on, and we don't want to ship the Developer's Guide
to end-users. However, we don't want to completely duplicate
conf.py for each manual, and we'd like to continue to
support "build all docs in one run" for third-party sites
like readthedocs.org.

Make the top-level conf.py support two usage forms:
 (1) as a common config file which is included by the conf.py
 for each of QEMU's manuals: in this case sphinx-build is run
 multiple times, once per subdirectory.
 (2) as a top level conf file which will result in building all
 the manuals into a single document: in this case sphinx-build is
 run once, on the top-level docs directory.

Provide per-manual conf.py files and top level pages for
our first two manuals:
 * QEMU Developer's Guide (docs/devel)
 * QEMU System Emulation Management and Interoperability Guide
   (docs/interop)
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Message-id: 20190305172139.32662-9-peter.maydell@linaro.org
Message-id: 20190228145624.24885-9-peter.maydell@linaro.org

sphinx-build complains about using :option: to mark up option
flags that it doesn't know about (because they were not defined
using the "option::" directive):
docs/pr-manager.rst:68: WARNING: unknown option: -d

Suppress these warnings. This way we get the semantic markup
of the option flag but no cross-referencing hyperlink.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-8-peter.maydell@linaro.org
Message-id: 20190228145624.24885-8-peter.maydell@linaro.org

Sphinx defaults to including all the rST source files
in the HTML build and making each HTML page link to the
source file. Disable that.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-7-peter.maydell@linaro.org
Message-id: 20190228145624.24885-7-peter.maydell@linaro.org

Add the 'navigation' bar to the sidebar, which for some
reason is not enabled by default. Remove 'relations', which
is effectively disabled anyway and isn't useful for us.

This requires that we mandate having at least Sphinx 1.3,
where the theme was added.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Message-id: 20190305172139.32662-6-peter.maydell@linaro.org
Message-id: 20190228145624.24885-6-peter.maydell@linaro.org

We don't yet have any custom static files, so disable this
config file setting to avoid a warning from sphinx about
not being able to find the directory.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-5-peter.maydell@linaro.org
Message-id: 20190228145624.24885-5-peter.maydell@linaro.org

Commit the initial Sphinx conf.py and skeleton index.rst as
generated with sphinx-quickstart. We'll update these to
add QEMU-specific tweaks in subsequent commits.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Acked-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20190305172139.32662-4-peter.maydell@linaro.org
Message-id: 20190228145624.24885-4-peter.maydell@linaro.org

Convert the memory API documentation from plain text
to restructured text format.

This is a very minimal conversion: all I had to change
was to mark up the ASCII art parts as Sphinx expects
for 'literal blocks', and fix up the bulleted lists
(Sphinx expects no leading space before the bullet, and
wants a blank line before after any list).
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Reviewed-by: NCleber Rosa <crosa@redhat.com>
Message-id: 20190305172139.32662-3-peter.maydell@linaro.org
Message-id: 20190228145624.24885-3-peter.maydell@linaro.org

sphinx-build complains:

docs/cpu-hotplug.rst:67: ERROR: Unexpected indentation.
docs/cpu-hotplug.rst:69: ERROR: Unexpected indentation.
docs/cpu-hotplug.rst:74: WARNING: Block quote ends without a blank line; unexpected unindent.
docs/cpu-hotplug.rst:75: WARNING: Block quote ends without a blank line; unexpected unindent.
docs/cpu-hotplug.rst:76: SEVERE: Unexpected section title.

}
{
docs/cpu-hotplug.rst:78: WARNING: Block quote ends without a blank line; unexpected unindent.

These are the result of not indicating one of the literal
blocks by finishing the preceding paragraph with the "::" marker.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAlex Bennée <alex.bennee@linaro.org>
Reviewed-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: NAleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Reviewed-by: NCleber Rosa <crosa@redhat.com>
Message-id: 20190305172139.32662-2-peter.maydell@linaro.org
Message-id: 20190228145624.24885-2-peter.maydell@linaro.org

usb: mtp fixes, guest-reset switch for usb-host.

# gpg: Signature made Thu 07 Mar 2019 09:53:55 GMT
# gpg:                using RSA key 4CB6D8EED3E87138
# gpg: Good signature from "Gerd Hoffmann (work) <kraxel@redhat.com>" [full]
# gpg:                 aka "Gerd Hoffmann <gerd@kraxel.org>" [full]
# gpg:                 aka "Gerd Hoffmann (private) <kraxel@gmail.com>" [full]
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/usb-20190307-pull-request:
  Introduce new "no_guest_reset" parameter for usb-host device
  usb-mtp: prevent null dereference while deleting objects
  usb-mtp: fix some usb_mtp_write_data return paths
  usb-mtp: return incomplete transfer on a lstat failure
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>